Summary
A Floating IP is a public IP address managed independently of a specific instance, allowing operators to move external reachability between instances without changing DNS records or reconfiguring clients.
What is a Floating IP?
In OpenStack and similar cloud platforms, every virtual machine receives a private IP address from its subnet via DHCP. A Floating IP is an additional public address from an external pool that can be associated with an instance's private port on demand. Under the hood, NAT is used to map incoming traffic from the Floating IP to the instance's private address.
The key advantage of Floating IPs over static public addresses is portability. If an instance fails, its Floating IP can be reassigned to a replacement instance within seconds, minimising downtime without requiring DNS changes or client reconfiguration.
Floating IPs are also a cost control mechanism: they can be disassociated when not needed and returned to the pool, so tenants only pay for public addresses that are actively in use.
Why is Floating IP relevant?
- High availability: Enables fast failover by reassigning public addresses without DNS propagation delays
- Flexibility: Decouples public reachability from individual instance lifecycles
- Cost efficiency: Public addresses can be released when not needed, reducing allocation costs